Correspondence between definitions and procedures: A reply to Stokes, Osnes, and Guevremont

Stokes, Osnes, and Guevremont's (1987) implicit definition of correspondence classes appears close to ours (Matthews, Shimoff, & Catania, 1987). Their definition, however, is fundamentally procedural and thus may have to be modified as experimental methodologies are refined. The advantage of our contingency-space analysis is that it is independent of specific procedures and focuses attention on problems inherent in some procedural definitions. Specifically, a contingency-space analysis addresses the issue of distinguishing specific instances from classes and reminds us that correspondence can be identified as a class only on the basis of observing a population of opportunities for say/do sequences in which the subject sometimes does not say.     

Intermittent reinforcement of a continuous response

Konorski showed that when a go/no-go procedure was used, sound quality discriminations were rapidly acquired and sound location discriminations were slowly acquired. These findings have been interpreted as a general constraint on the acquisition of auditory discriminations (quality-location effect). However, experiments carried out within an evolutionary framework (Harrison, 1984) have shown that the rate of acquisition of sound location discriminations varies widely as a function of the inclusion or exclusion of naturalistic features. These data suggest that Konorski's findings were a function of the special conditions of the experiments. The first purpose of the present experiments was to assess whether rats showed the effects noted by Konorski when studied under similar conditions. The second purpose was to study the effect of manipulating two natural features (novelty and stimulus-response adjacency) to assess whether the acquisition rates of quality and location discriminations could be greatly modified or made approximately equal, or both. When a go/no-go procedure was used and the other conditions were similar to those of Konorski, rats acquired a quality discrimination but did not acquire a location discrimination. However, when the S+ or S- were presented through a closely adjacent speaker, the sound location discrimination was acquired as rapidly as the quality discrimination. Finally, preexposing the animal to either S+ or S- retarded the rate of or prevented the acquisition of the quality discrimination. The experiments showed that the quality-location effect was determined primarily by the conditions used in Konorski's experiments, and that the effect is not a general constraint on learning.     

Rapid determinations of preference in multiple concurrent-chain schedules.

With concurrent chains arranged for a pigeon's key pecks, pecks on two concurrently available initial-link keys (left and right) respectively produce separately operating terminal links (A and B). Preferences for terminal link A over terminal link B are usually calculated as deviations of relative initial-link response rates (left divided by total pecks) from those during baseline conditions, when A equals B. Baseline preferences, however, are often variable and typically are determined indirectly (e.g., with unequal A and B, reversing left-right assignments of A and B over sessions and estimating the baseline from differences between the relative rates generated). Multiple concurrent-chain schedules, with components each consisting of a pair of concurrent chains, speed the determination of preferences by arranging A and B and their reversal within sessions. In two experiments illustrating the feasibility of this procedure, one component operated with circles projected on initial-link keys and the other with pluses; when left and right initial-link pecks respectively produced terminal links A and B in one component, they produced B and A in the other. Even as the baselines fluctuated, preference was observable within sessions as the difference between relative initial-link response rates in the two components. The first experiment demonstrated the rapid development of preferences when terminal links A and B consisted of fixed-interval 15-s and 30-s schedules. The second demonstrated the sensitivity of the procedure to preference for a fixed-interval 30-s schedule operating for pecks on either of two keys (free choice) over its operating for pecks on only a single key (forced choice).     

Similarities in the rate-altering effects of white noise and cocaine.

The effects of white noise and cocaine on squirrel monkeys' fixed-interval responding were compared to determine whether the presentation of an exteroceptive stimulus could produce rate-altering effects of the type typically observed following drug administration. To investigate the relationship between control response rate and response rate in the presence of drug or noise, the monkeys were trained under a fixed-interval 300-s stimulus-shock termination schedule in order to generate a wide range of local response rates. A light illuminated the experimental chamber during the interval and, after 300 s elapsed, a lever press during a 3-s period terminated the light and precluded the occurrence of a harmless electrical stimulus that otherwise was delivered at the end of the 3-s period. Each interval was followed by a 30-s timeout during which the chamber was darkened and responses had no consequences. Following intramuscular administration of cocaine, different rates of responding characteristic of control performance converged toward a common rate and, at an appropriately high dose, response rate during the fixed interval became more uniform. When white noise was presented continuously during a given session, different response rates also converged toward a common rate and, at an appropriate intensity, response rate became more uniform. Interactions were obtained when cocaine and white noise were presented together, indicating the possibility of a common behavioral mechanism of action. The results suggest that rate-altering drug effects may be, in part, a result of the ability of drugs to produce nonspecific stimulus effects similar to those observed for exteroceptive stimuli.     

Reinforcement probability and ordinal position of response in fixed-interval schedules

Four rats pressed levers and received food pellets under fixed-interval reinforcement schedules of 20, 60, and 180 seconds. The number of responses in each interval was recorded. From these data, the probability of reinforcement was determined as a function of response count. These functions were generally increasing. This finding is consistent with previous suggestions that increasing response rates within fixed intervals may be a function of response count in addition to or instead of elapsed or remaining time.     

Delayed reinforcement and delayed choice in symbolic matching to sample: Effects on stimulus discriminability

Six pigeons were trained to peck a red side key when the brighter of two white lights (S₁) had been presented on the center key, and to peck a green side key when the dimmer of two white lights (S₂) had been presented on the center key. Equal frequencies of reinforcers were provided for the two types of correct choice. Incorrect choices, red side-key pecks following S₂ presentations and green side-key pecks following S₁ presentations, resulted in blackout. With 0-s delay between choice and reinforcement, the delay between sample presentation and choice was varied from 0 to 20 s. Then, with 0-s delay between sample presentation and choice, the delay between choice and reinforcement was varied from 0 to 20 s. Both types of delay resulted in decreased discriminability (defined in terms of a signal-detection analysis) of the center-key stimuli, but delayed choice had more effect on discriminability than did delayed reinforcement. These data are consistent with the view that the two kinds of delay operate differently. The effect of a sample-choice delay may result from a degradation of the conditional discriminative stimuli during the delay; the effect of a choice-reinforcer delay may result from a decrement in control by differential reinforcement.     

The effect of signaled reinforcement availability on concurrent performances in humans

During Phase I, three female human subjects pressed a button for monetary reinforcement in two-component concurrent variable-interval schedules. Five different reinforcement frequencies were used in component A, whereas the reinforcement frequency in component B was held constant. Absolute rates of responding conformed to equations proposed by Herrnstein to describe concurent performances, and the ratios of the response rates and the times spent in the two components conformed to the matching law. During Phase II, the availability of reinforcement in component A was signaled by the illumination of a lamp. This resulted in suppression of response rates in component A and elevation of response rates in component B, these changes being reflected in a distortion of the matching relationship which took the form of a bias in favor of component B.     

Chained schedules of avoidance: Reinforcement within and by avoidance situations

Four rats were exposed to chained schedules with variable-cycle avoidance in both links. Responding in the initial link cancelled shocks scheduled once per minute and, according to a conjoint fixed-ratio schedule, produced a terminal link where scheduled shock rates varied from 0 to 8 shocks per minute in different conditions of the experiment. Response rates in the terminal link increased as a function of the scheduled shock rate. Response rates in the initial link, on the other hand, decreased as a function of the shock rate actually received (rather than scheduled) in the terminal link. While consistent with other studies of aversive control, these results differ from those obtained in chained schedules of positive reinforcement in that increases in reinforcement within the terminal link of the chain did not systematically increase the reinforcing value of that link.